Legacy and Mainframe Migration: Finally the Rule Rather Than the Exception
By Chad Hersh, Senior Analyst, Celent
Insurance carriers have been making efforts to deal with their legacy systems for longer than anyone cares to remember. Though the industry was among the earliest adopters of technology, it has done less to stay current than many other sectors – perhaps because of this legacy burden.
A History of Excuses
The most common reason given for resisting legacy modernization – and the one that often stops arguments in their tracks at many carriers – is, “Our system still works fine.” However, “fine” often means that the system runs reliably and performs its core functions well, but does not mean that the system is meeting all of the business needs of a carrier.
Another frequently heard excuse is that, “It’s too expensive/hard/resource intensive/time consuming to replace” one’s legacy systems. However, a typical carrier spends between 50% and 70% of its budget maintaining IT systems and infrastructure rather than working on new projects. This has a devastating effect on a carrier’s innovation and ability to compete effectively against carriers with less extensive legacy burdens.
The excuse that “the system is poorly documented and would be difficult or impossible to recreate” is a big, red warning flag that your system is likely hampering any efforts you might make to improve your business processes. If your system is poorly documented to the point that it could not be recreated in a new system, this typically means that changing existing business processes within the system is likely to be exceedingly difficult, because the logic is probably hard-coded and poorly documented.
Often the rationale that “no packages meet our needs” is simply an indication that years of inflexible business processes have led a carrier to believe that it must approach its business in a particular way. Replacing core systems is an excellent opportunity to revisit certain processes. Additionally, a line of business or a piece of functionality that does not exist in a modern package today can often be added quickly and easily.
Another common excuse is that a carrier “needs the scalability/reliability of the mainframe.” While just a few years ago this was a valid excuse for most carriers, today’s commodity hardware, when combined with clustering technology or grid technology, can not only meet the reliability and scalability needs of all but the largest carriers, but can usually do so for a fraction of the total cost of mainframe ownership.
Finally, the argument that “our mainframe is fully depreciated” is perhaps the worst argument of all. Migrating from the mainframe and to a server-based platform can save tens or hundreds of thousands of dollars per year, and in many cases several million dollars per year. These savings come from eliminating the often incredibly costly license fees for mainframe system management tools, eliminating the need for employees with rare skill sets that are only getting rarer, eliminating the cost of replacement parts, etc.
Overcoming the Excuses
Many technologies are converging to finally make migrating from legacy systems hardware realistic and cost-effective. Though past promises proved empty, an array of real technologies has emerged to make migrating away from legacy systems a viable, attractive option. Virtualization and grids can give commodity hardware tremendous computing power. Legacy migration tools and specialized vendors make getting from legacy to modern less painful. Business rules engines and BPM solutions make rewriting hard-coded business rules simpler and more flexible. Perhaps most importantly, Web services and SOA make migrating one piece at a time far more cost effective than in the past.
This combination of factors creates a powerful argument for migrating now. Though still not a simple process, doing so has tremendous potential for cost-savings and risk reduction. Any carrier not already exploring their migration options should at least be reviewing the landscape of available options.